NUI Maynooth Eprint Archive
Not a member yet
18159 research outputs found
Sort by
More efficient Geospatial ML modelling techniques for identifying man-made features in Aerial Ortho-imagery
No full textDeep learning techniques are used to achieve state-of-art accuracy in semantic segmentation on
aerial ortho-imagery datasets. These algorithms are known to be efficient in terms of accuracy but at the
expense of computational power required for training and subsequent inference operations. In this paper
we strive to achieve a comparable performance but with lower floating point operations per second
(FLOPS) and less training time. With this in mind, we chose to evaluate the EfficientNet-B0 network
configured with 5.3 millions parameters and 0.39 billion FLOPS as a feature extractor operating inside a
U-net architecture, achieving accuracy levels (mean F1 score of 0.869) comparable to a state-of-the-art
deep learning architecture (U-net with Resnet50 as backbone) configured with 25.6 million parameters
and 4.1 billion FLOPS which achieved a mean F1 score of 0.87. These promising results demonstrate that
employing EfficientNet as the feature extractor in semantic segmentation on aerial ortho-imagery can be
an effective strategy, in achieving higher performance results in terms of computational power, especially
when running these networks on the edge
Economics of Social Issues
No full textThis thesis deals with the analysis of various social issues from an economic perspective. Chapter one looks at the issue of gun crime from a game theoretical perspective, analysing the differing effects of concealed carry of handguns and open carry of handguns on the number of crimes committed. It is found that open carry only laws lead to fewer crimes than concealed carry only laws. In states where both types of carry are permitted, attitudes in favour of firearms are shown to increase carry rates and lower crime. Chapter two deals with the topic of suicide and mental health issues, examining empirically the effect of various economic indicators, such as Gross Domestic Product (GDP), unemployment and working hours on the suicide rate and the depression rate. Ordinary Least Squares (OLS) regressions show that a higher GDP is associated with lower rates of both suicide and depression. Chapter three is a theoretical paper which analyses the effect of corporate donations on the policy position of political parties. It is found that with a small minority supporting donors’ positions, donations can have a large effect on policy positions. Under certain circumstances, parties can collude on policy position, at the expense of voters, to gain more donations
Creating new Program Proofs by Combining Abductive and Deductive Reasoning
No full textWe describe recent work on the Aris system that creates and
verifies new formal specifications for pre-existing source
code. We describe Aris in terms of the abductive reasoning
system that suggest possible specifications and then uses an
existing deductive verifier to evaluate these creations. This
paper focuses on the abduction system that creates new formal specifications by leveraging a small set of inspiring artefacts to augment a subset of candidate problems. This employs knowledge graphs to represent the raw data (i.e., source
code), discovering latent similarities between graphs using a
graph-matching process. Results are presented for the C# programming language with novel creations and its sister language called Code Contracts. We outline ampliative creativity, whereby newly created artefacts drive subsequent creative episodes beyond the initially perceived limitations. We
also outline some recent work towards transferring specifications between the C# and Java programming languages
Conditioning ensemble streamflow prediction with the North Atlantic Oscillation improves skill at longer lead times
No full textSkilful hydrological forecasts can benefit decision-making in water resources management and other water-related sectors that require long-term planning. In Ireland, no such service exists to deliver forecasts at the catchment scale. In order to understand the potential for hydrological forecasting in Ireland, we benchmark the skill of ensemble streamflow prediction (ESP) for a diverse sample of 46 catchments using the GR4J (Génie Rural à 4 paramètres Journalier) hydrological model. Skill is evaluated within a 52-year hindcast study design over lead times of 1 d to 12 months for each of the 12 initialisation months, January to December. Our results show that ESP is skilful against a probabilistic climatology benchmark in the majority of catchments up to several months ahead. However, the level of skill was strongly dependent on lead time, initialisation month, and individual catchment location and storage properties. Mean ESP skill was found to decay rapidly as a function of lead time, with a continuous ranked probability skill score (CRPSS) of 0.8 (1 d), 0.32 (2-week), 0.18 (1-month), 0.05 (3-month), and 0.01 (12-month). Forecasts were generally more skilful when initialised in summer than other seasons. A strong correlation (Ï=0.94) was observed between forecast skill and catchment storage capacity (baseflow index), with the most skilful regions, the Midlands and the East, being those where slowly responding, high-storage catchments are located. Forecast reliability and discrimination were also assessed with respect to low- and high-flow events. In addition to our benchmarking experiment, we conditioned ESP with the winter North Atlantic Oscillation (NAO) using adjusted hindcasts from the Met Office's Global Seasonal Forecasting System version 5. We found gains in winter forecast skill (CRPSS) of 7 %-18 % were possible over lead times of 1 to 3 months and that improved reliability and discrimination make NAO-conditioned ESP particularly effective at forecasting dry winters, a critical season for water resources management. We conclude that ESP is skilful in a number of different contexts and thus should be operationalised in Ireland given its potential benefits for water managers and other stakeholders
Impact of intelligent control algorithms on demand response flexibility and thermal comfort in a smart grid ready residential building
No full textThe present paper investigates the impact of advanced control algorithms on harnessing building energy flexibility in a smart-grid ready full-electric residential building. The impact on thermal comfort is also analysed. The building is located in Ireland and is equipped with a geothermal heat pump and a thermal energy storage system. Two Energy Management systems, based on rule-based and intelligent optimisation algorithm approaches, are developed which use real-time building smart meter and weather data. This data is utilised by various dynamic flexibility metrics within the respective control algorithms. Different time of use tariffs, based on data from the Irish Commission for Energy Regulation and structured on the basis of peak, off-peak and night periods, are also used. Results show that energy cost reductions of up to 21% and 43% can be achieved by the rule-based and intelligent algorithm, respectively, without compromising the thermal comfort within the building. Moreover, total shifting and forcing flexibility potential of up to 34 and 54 kWh, respectively, based on the month of January, can be achieved by the adoption of the intelligent control algorithm
Perceptions of Change in the Natural Environment produced by the First Wave of the COVID-19 Pandemic across Three European countries. Results from the GreenCOVID study
No full textAlthough different studies have evaluated the positive impacts of the COVID-19 pandemic and lockdown mea-
sures on reducing noise pollution and traffic levels and improving air quality, how populations have perceived
such changes in the natural environment has not been adequately evaluated. The present study provides a more
in-depth exploration of human population perception of enhanced natural exposure (to animal life and nature
sounds) and reduced harmful exposure (by improved air quality and reduced traffic volume) as a result of the
COVID-19 pandemic lockdown. The data is drawn from 3,109 unselected adults who participated in the
Green COVID survey from April to July 2020 in England, Ireland, and Spain. The findings suggest that the positive
impacts to the natural environment as a result of the lockdown have been better received by the population in
Spain and Ireland, in comparison to England. Participants who resided in urban areas had better perceived
improvements in nature sounds, air quality, and traffic volume compared to those in rural areas. Older pop-
ulations and those with lower smoking and alcohol consumption were found to perceive this improvement the
most. Furthermore, the greater perception of improvements in environmental elements was also associated with
better self-perceived health and improved wellbeing. In the binary logistic regression, living in Ireland or Spain,
urban areas, female gender, older age, and good overall wellbeing were associated with a greater perception of
improvements in the natural environment, while the factors most associated with a greater perception of reduced
harmful exposure were living in Spain, had a good self-perceived health status and older age
Modeling and Analysis of Dynamic Charging for EVs: A Stochastic Geometry Approach
No full textWith the increasing demand for greener and more energy efficient transportation solutions, electric vehicles (EVs) have emerged to be the future of transportation across the globe. However, currently, one of the biggest bottlenecks of EVs is the battery. Small batteries limit the EVs driving range, while big batteries are expensive and not environmentally friendly. One potential solution to this challenge is the deployment of charging roads, i.e., dynamic wireless charging systems installed under the roads that enable EVs to be charged while driving. In this paper, we use tools from stochastic geometry to establish a framework that enables evaluating the performance of charging roads deployment in metropolitan cities. We first present the course of actions that a driver should take when driving from a random source to a random destination in order to maximize dynamic charging during the trip. Next, we analyze the distribution of the distance to the nearest charging road. This distribution is vital for studying multiple performance metrics such as the trip efficiency, which we define as the fraction of the total trip spent on charging roads. Next, we derive the probability that a given trip passes through at least one charging road. The derived probability distributions can be used to assist urban planners and policy makers in designing the deployment plans of dynamic wireless charging systems. In addition, they can also be used by drivers and automobile manufacturers in choosing the best driving routes given the road conditions and level of energy of EV battery
Book Review: Practical strategies in geriatric mental health: Cases and approaches
No full textThe abstract is included in the tex
Do publication activities of academic institutions benefit from formal collaborations with firms?
No full textWhile the existing literature has focused predominantly on how firms can benefit from collaborations with academic institutions such as universities and research institutions, this study explores whether the proportion of (formal) collaborations with different types of firm partners in strategic R&D alliances is associated with publications originating in academic institutions. The empirical analysis is based on a unique dataset of publications in pharmaceutical cancer research. The results suggest that the share of collaborations with industry partners has an inverted u-shaped relationship with the reputation of the journal in which an article originating in an academic institution is published. The share of alliances with pharmaceutical firms shows a similar inverted u-shaped pattern, suggesting that research originating in academic institutions can only benefit from alliances with pharmaceutical firms through resource inflows up to a threshold
Longitudinal changes in psychological distress in the UK from 2019 to September 2020 during the COVID-19 pandemic: Evidence from a large nationally representative study
No full textIn a large (n=10918), national, longitudinal probability-based sample of UK adults the prevalence of clinically
significant psychological distress rose from pre-pandemic levels of 20.8% in 2019 to 29.5% in April 2020 and then
declined significantly to pre-pandemic levels by September (20.8%). Longitudinal analyses showed that all demographic groups examined (age, sex, race/ethnicity, income) experienced increases in distress after the onset of
the pandemic followed by significant decreases. By September 2020 distress levels were indistinguishable from
pre-pandemic levels for all groups. This recovery may reflect the influence of the easing of restrictions and
psychological adaptation to the demands of the pandemic